| Summary: | Breast cancer imaging currently improved for the past two decades due to digitalization and computer development. Studies have shown that early cancer diagnosis can improve the life expectancy of patients. The most effective and popular technique for detecting the early modalities or other abnormalities in human breast tissue is digital mammography. Noise is a big problem in mammogram images that must be tackled for mammogram research. Blurring artifact normally occurs in the reconstructed projection image as its side effect. This paper focus on how to enhance the mammogram images that have been corrupted with random value impulse noise. The Modified Hybrid Median Filter (MHMF) is proposed to be used in this project based on its best performance eliminating the RVIN shown in previous research. The filter consists of two stages i.e. the noise detection phase and the filtering phase. Some modification needs to be done to the MHMF as the nature of digital mammogram image consist of huge black pixel. Therefore, a pre-processing step is needed before the MHMF can be used. The result shows that the integration of MHMF with the pre-processing phase give a better performance, outperform the performance of MHMF alone. The outcome of this project improved the quality of the corrupted mammogram images and could assist radiologist in diagnosing their patient. © 2022 American Institute of Physics Inc.. All rights reserved.
|
|---|